AI 中文总结
针对无范例半监督类增量学习的灾难性遗忘与伪标签不可靠问题，提出MAGIC框架，通过软加权几何校准与几何结构对齐组件，在多数据集低标签场景下提升了平均增量准确率。
AI 中文摘要
半监督类增量学习（SSCIL）是神经网络面临的严峻挑战，在无范例设置中难度最大，该设置下无法存储任何过往数据。现有方法因特征漂移会发生灾难性遗忘，且随着标签空间扩大，其伪标签的可靠性日益降低。本文提出MAGIC（流形锚定与几何增量校准）框架，可在不存储范例的情况下稳定可塑性。MAGIC的设计围绕两个组件展开：第一个是软加权几何校准（SWGC），它在学习器的可塑性特征空间上使用基于图的标签传播，对在冻结骨干网络上计算的类别均值和方差进行加权与校准；从这些校准后的高斯分布中，我们采样代表过往任务数据的伪特征。第二个是几何结构对齐（GSA）目标，它通过匹配学生头和教师头的关系结构，并将特征锚定与固定分类器原型对齐，来保留表示拓扑结构，锁定特征空间的方向。这些约束共同防止适配器发生漂移，使得随着新类别加入，类别间的几何关系保持稳定。我们使用冻结的ResNet-18骨干网络和可学习的可塑性适配器实现MAGIC。在CIFAR-100、CUB-200和ImageNet-R数据集上，当标签比例为1%、5%和10%时，MAGIC相较于配备FixMatch的多数监督CIL方法及原生SSCIL基线，提升了平均增量准确率；在细粒度、低标签设置中增益最大，该设置下置信度阈值法的失效最为明显。
英文摘要
Semi-supervised Class Incremental Learning (SSCIL) is a severe challenge for neural networks, and it is hardest in the exemplar-free setting where no past data may be stored. Existing methods forget catastrophically due to feature drift, and their pseudo-labels become increasingly unreliable as the label space grows. In this paper, we propose MAGIC (Manifold Anchoring and Geometric Incremental Calibration), a framework that stabilizes plasticity without storing exemplars. MAGIC's design centers on two components. The first is Soft-Weighted Geometry Calibration (SWGC), which uses graph-based label propagation on the learner's plastic feature space to weight and calibrate class means and variances computed on the frozen backbone; from these calibrated Gaussians, we sample phantom features that stand in for data from previous tasks. The second is a Geometric Structural Alignment (GSA) objective that preserves representation topology by matching the relational structure of student and teacher heads and aligning feature anchors with the fixed classifier prototypes, locking the orientation of the feature space. Together, these constraints keep the adapter from drifting, so geometric relations between classes remain stable as new classes arrive. We implement MAGIC with a frozen ResNet-18 backbone and a learnable plastic adapter. Across CIFAR-100, CUB-200, and ImageNet-R, at label ratios of 1%, 5%, and 10%, MAGIC improves average incremental accuracy over most of the supervised CIL methods equipped with FixMatch and native SSCIL baselines; the largest gains occur in the fine-grained, low-label setting, where confidence thresholding fails most clearly.
